Part II: GFS Model Verification for 28 October 2003 - Medium Range Forecast Exercise for Tallahassee, Florida

Medium Range Forecast Exercise for Tallahassee, Florida: A Prototype for Forecasting Along the Gulf Coast in Fall

The forecaster will view model verifications for 28 October 2003 and then make appropriate modifications to the day 5 forecast made in Part I (Section 4) of this exercise. These model initializations for 28 October 2003 would not have been available, of course, in real time to the forecaster on 23 October 2003. The initializations are presented here as verifications of the 5-day NWP forecast from 23 October 2003 for the instructional purposes of this exercise. The forecaster is given a "second chance", which would not be available in real time, to modify the forecast made from 23 October 2003 model runs, using the model verifications presented here for 28 October 2003.

Part II, GFS Model Verification for 28 October 2003

Examine Figure 5 which shows the GFS 500 hPa height and absolute vorticity in the model's initialization for 0000 UTC 28 October 2003.

Figure 5: GFS 500 hPa height (orange) and absolute vorticity (green) in the initialization for 0000 UTC 28 October 2003.

Compare this initialization to the GFS 120-hour 500 hPa height and absolute vorticity forecast verifying at the same time, shown in Figure 6. This figure was presented in Part I, Section 2 as Figure 4.

Figure 6: GFS 120-hour 500 hPa height (orange) and absolute vorticity (green) forecast verifying at 0000 UTC 28 October 2003.

For convenience in viewing the forecast and verifying analysis the two are overlain in Figure 7. Note that in this graphic the initialization and 120-hour forecast verification are for 1200 UTC 28 October 2003, 12 hours later than for Figures 5 and 6.

Figure 7: GFS 120-hour 500 hPa height forecast valid at 1200 UTC 28 October 2003 (red) superimposed on the verifying initialization for 1200 UTC 28 October 2003 (blue).

Now, consider the following question.

Question A, Part II

Discussion

The verification will necessitate major changes to the forecast. The forecast of cold, dry, clear, and precipitation-free must be replaced with warmer temperatures, greater moisture, more cloudiness and the likelihood of precipitation. The cold, dry upper air vortex centered over Indiana with its associated westerly flow over North Florida (Fig. 6) did not materialize. What we now see is a significant trough over Texas giving North Florida southwesterly flow aloft. This would indicate significantly warmer temperatures and greater moisture influx from the Gulf of Mexico. There is enough PVA ahead of this trough (Fig. 5) to expect that it will propagate eastward toward the forecast area bringing cloudiness and precipitation. If, indeed, the residual moisture from Tropical Depression Patricia is incorporated into this trough, the rainfall for TLH could be heavy.

With a warmer air mass accompanied by greater moisture and cloudiness, the minimum temperature should be raised from MOS's value of 38 and the ZFP value in the lower 40's. It should be higher than the climatological mean value of 52 as well, considering the southwesterly flow aloft. The maximum temperature forecast for MOS (64) and the ZFP (upper 60's) may not need much adjustment. This is a case of getting it right for the wrong reason. The MOS and ZFP forecasts were considerably cooler than the climatological mean of 78 for the date because of the expected cool, dry air mass. The maximum temperature will likely be well below the climatological mean, but this will be due to the expected cloudiness and precipitation.

A reasonable revised forecast is: cloudy with rain possibly heavy at times, low in the lower 60's, high in the upper 60's.

		a) The orientation of the long wave trough in the eastern U.S.
		b) The location and intensity of vorticity centers in the longwave trough in the eastern U. S.

Medium Range Forecast Exercise for Tallahassee, Florida: A Prototype for Forecasting Along the Gulf Coast in Fall

Question A, Part II

Question B, Part II